Typically the powder stabilizers are more potent than their liquid counterparts. This is probably due to the fact that the liquid stabilizers must be soluble in a liquid carrier and this solubility is limited. The entire formulation would require an adjustment after a switch from powder to liquid, as these materials have vastly different lubricity and fusion characteristics.
The details of this adjustment are beyond the scope of this forum and would require a total understanding of your formulation and process. I'd look for a powder Ba/Zn stabilizer as a booster to the current liquid or replace the Ba/Zn liquid completely with a Ba/Zn powder. The entire formulation will need to be adjusted for lubricity and possibly will need a process aide (Acryloid K-120) addition, as well for smooth degradation free extrusion. These recommendations are given with the undersanding that nothing can be done w/out testing, observation, and patient re-formulation work.
I would expect about 15% recycling of your re-processed scrap and trimmings to be about right, given the limitations of your current formulation. If you can improve the batch-to-batch stability and consistency of your compound, then you may be able to recycle a bit more scrap and trim.
To your former question about completely removing and dismantling the head during a clean out, this is completely necessary with f-PVC. If you leave a small spot of degraded PVC in the die or adapter and restart the line, that residual PVC will catalyze degradation and burning will start in much less than 12 hours.
As to your resin question, the PVC resin is probably not the root cause of your problem. I’d suspect the formulation is not properly stabilized, first. And, next I’d concentrate on your mixing and compounding process to standardize level out the process between batches.